Using Probabilistic Logic Programming to Find Patterns

Theofrastos Mantadelis, Ricardo Rocha, Jorge Oliveira and Miguel Tavares Coimbra

October 2015

---

Abstract

This short paper, briefly presents the probabilistic logic programming language ProbLog and the system MetaProbLog. We present an example Hidden Markov Model to illustrate the three main tasks of the system. Furthermore, we mention some of the existing ProbLog applications which are used to find connections/patterns in relational databases. Finally, we present an application that uses MetaProbLog for phonocardiogram classification which is used in order to diagnose heart diseases.

Bibtex

@InProceedings{mantadelis-recpad15,
  author =    {T. Mantadelis and R. Rocha and J. Oliveira and M. T. Coimbra},
  title =     {{Using Probabilistic Logic Programming to Find Patterns}},
  booktitle = {Proceedings of the 21st Portuguese Conference on Pattern Recognition (RecPad 2015)},
  editor =    {J. Rodrigues and P. Cardoso and R. Lam and M. Figueiredo},
  month =     {October},
  year =      {2015},
  address =   {Faro, Portugal},
}

Download Paper

PDF file

Download Poster

PDF file